基于SOM神经网络的内燃机故障诊断模型

Fault Diagnosis of Internal Combustion Engine Based on SOM Neural Network

本文在总结自组织映射(Self-OrganizingMap,SOM)神经网络的工作原理及实现过程的基础上,分析了内燃机的典型性能指标与故障类型之间的关系,并以某型号内燃机的振动信号为样本,利用经迭代训练后的网络结构对其进行检测,以此判断设备的工作状态。基于SOM神经网络的内燃机故障诊断模型,具有较好的诊断效果。

Based on the summary of the working principle and implementation process of Self-Organizing Map(SOM) neural network, this paper analyzes the relationship between typical performance indexes and fault types of internal combustion engines, and uses the vibration signal of a model of internal combustion engine as a sample to detect it using the network structure after iterative training, so as to determine the working status of the equipment. The fault diagnosis model of internal combustion engine based on SOM neural network has a good diagnosis effect.